Quasi-realism is an anti-realist non-cognitive meta-ethical theory because it argues that no moral statements are fundamentally descriptive; they don’t describe a moral reality. Rather moral statements are statements that ‘ project our sentiments’ (Blackburn in Miller 2003: 52) on to actions. The meaning of moral statements is not what they refer to (because quasi-realism is anti-realist, there is no moral property to which moral statements refer) rather the meaning of such statements is the projection of our emotions and values (which are non-descriptive) on to actions. But quasi-realism is to be distinguished from other non-cognitive theories by its attempt to explain and justify the use of moral language which seems to be cognitive (e. g. which seems to have the feature of describing a moral reality, and not merely expressing our sentiments) even though it isn’t. Quasi-realism is the position that we express moral statements as if they are beliefs even though they are not. One of the main motivations for quasi-realism is the Frege-Geach problem which poses a serious problem for non-cognitivist theories. I will argue that quasi-realism is successful in dealing with the Frege-Geach problem and can overcome one of the most problematic objections to it.
To understand quasi-realism it is necessary to understand the distinction between cognitive and non-cognitive metaethical theories of moral discourse. According to cognitive theories a moral statement expresses a belief whilst according to non-cognitive theories moral statements do not express beliefs but rather some other kind of mental state. The best way to characterise the difference between a belief and some other mental state that is not a belief is in terms of description. A mental state that represents something in the world can be assessed for truth or falsity because it purports to describe how something is in the world and only if that something in the world is as the mental state describes it is that description true. For example we have the non-moral belief that the sky is blue which is an attempt at describing something in the world. Likewise moral statements, according to the cognitivist, are expressions of mental states which attempt to describe how the world is. For example, ‘ Genocide is wrong’ is a moral statement which expresses a mental state which describes something in the world i. e. the event of genocide has the moral property of wrongness. The non-cognitivist disagrees with cognitivist that this statement is genuinely descriptive because for him moral statements do not express descriptive mental states but rather non-descriptive mental states. There are many mental states that we would call non-descriptive such as the mental state of happiness which does not attempt to describe the world in some way. It is important to note that description is not the same as intentionality even though description is intentional. That is, even though all descriptive mental states are intentional because all descriptive mental states are about something not every mental state that is about something also describes that something. For example the mental state of happiness may be about something e. g. we might be happy that our football team won, but that mental state does not describe something in the world. The difference between intentional descriptive mental states and intentional non-descriptive mental states is, roughly, that the former involves the use of concepts whilst the latter does not. When we are happy about something we are not using any concepts whilst when we have the belief that the sky is blue we are using the concepts of blue and sky. So the difference between cognitivist and non-cognitivist theories is that the former takes the meaning of moral statements as descriptions of the world whilst the latter take the meaning of moral statements to be the expression of non-descriptive mental states. For the non-cognitivist the meaning of a moral statement is just the emotion/value/sentiment that is expressed by that statement.
The main motivation for quasi-realism is the problem that all non-cogntivist theories faced. This is the Frege-Geach problem. The problem is best seen in the form of a typical moral argument.
Vivisection is wrong.
If vivisection is wrong then it is wrong for the government to fund vivisection.
Therefore, it is wrong for the government to fund vivisection.
The problem for the non-cognitivists comes when he tries to explain how this argument could be valid. The argument looks obviously valid when we see that it is simply a modus ponens argument with a conditional and the antecedent of the conditional guaranteeing the consequent. Unfortunately for the non-cognitivist he cannot say that this argument is valid. To see why it must be remembered that for the non-cognitivist the meaning of a moral statement is the non-descriptive mental state that it expresses. (1) expresses a non-descriptive mental state about vivisection (such as disapproval of it) and so the meaning of (1) might be equated to ‘ I disapprove of vivisection’ for the non-cognitivist. When (2) is considered, the same statement appears as the antecedent but the non-cognitivist would not give it the same meaning as (1) has; he would not equate ‘ vivisection is wrong’ with ‘ I disapprove of vivisection’. The reason is that someone who expresses (1) is expressing their non-descriptive mental state but someone who expresses (2) (necessarily expressing the antecedent of (2)) is not expressing the same non-descriptive mental state. He cannot be expressing his disapproval of vivisection in (2) because he hasn’t asserted that vivisection is wrong only that if vivisection is wrong then government funding of vivisection is wrong. (1) is asserted and so expresses disapproval but there is no assertion in (2) because the speaker is not judging vivisection to be wrong (because it is part of a conditional statement) so it seems that the speaker is not expressing the same non-descriptive mental state to the effect that he disapproves of vivisection. Because non-cognitivism ties the meaning of a statement to the non-descriptive mental state expressed it follows that as (1) and the antecedent of (2) do not express the same non-descriptive mental state then they don’t have the same meaning. The reason that the non-cognitivism entails the above argument as being invalid is now clear. For if (1) has a different meaning to the antecedent in (2) then the argument is guilty of the fallacy of equivocation. The fallacy of equivocation is using the same term (the same string of letters) in more than one sense (i. e. with different meanings). For example if I argued that
A piece of paper is light.
If something is light then it is not dark.
Therefore a piece of paper is not dark.
then I would be committing the fallacy of equivocation because I have used a term in more than one way. Similarly the non-cognitivist must maintain that the argument for government funding of vivisection being wrong commits a fallacy of equivocation because the phrase ‘ vivisection is wrong’ has a different meaning in (1) as it does in (2). Of course this is a problem for the non-cognitivist because this argument is seemingly valid. It seems very counter-intuitive to say that this argument is invalid, because it doesn’t seem to commit the fallacy of equivocation in the obvious way that the above argument does. Thus the challenge for the non-cognitivist is to explain how this argument can be valid.
For later purposes it is important to understand why the cognitivist does not face a similar problem in the Frege-Geach argument. The cognitivist understanding of the given argument doesn’t face the problem of equivocation because the meaning of the antecedent of (2) is the same as the meaning of (1). They have the same meaning because they both express the same descriptive mental state which is a description of the wrongness of vivisection. Whilst (1) asserts the truth of the wrongness of vivisection and (2) does not they still have the same content. This is unlike the non-cognitivist who must give the statements different meanings because for him only assertion is tied to a non-descriptive mental state and a statement which does not assert anything cannot be tied to a non-descriptive mental state and they must therefore have different meanings.
The quasi-realist attempts to respond to the Frege-Geach argument by introducing the notion of moral sensibility. Someone’s moral sensibility is all of their attitudes towards certain actions. For example someone may have the positive attitude towards somebody telling the truth and somebody else may have a negative attitude towards people telling lies. This fits in well with non-cognitivism because attitudes are non-descriptive mental states. Whilst we can have certain attitudes towards actions ‘ it is important that we can not only take up an attitude towards people’s actions but also towards their moral sensibilities’ (McNaughton 1988: 183). Blackburn reads conditionals such as (2) as being attitudes towards moral sensibilities. So (2) reads as an attitude of approval towards people who hold both an attitude of disapproval to vivisection and an attitude of disapproval to government funding of vivisection. To show how this solves the Frege-Geach problem it is necessary to use Blackburn’s own system of representing such attitudes. The two types of attitude (approval and disapproval) are represented by H! And B! Respectively with the action that is approved put in parentheses. So (1) would come out as B!(vivisection). Blackburn allows second order attitudes of approval and disapproval with the use of |…| which ‘ indicates that we are talking about the expression contained between the bars’ ( Grefenstette 2007: 3) so that disapproving of those who approve of genocide would be represented as B!(| H!(genocide)|). Finally Blackburn allows us to approve of those who approve/disapprove of two or more actions by inserting a semi-colon between the actions which are approved or disapproved. As we said that (2) is to be read as an approval of a moral sensibility which contained both the attitude of disapproval to vivisection and the attitude of disapproval to government funding of vivisection then (2) can be read as H!(| B!(Vivisection)|;| B!(Government funding of vivisection)|). So we can represent the argument for (3|) in terms of attitudes as:
B!(vivisection)
H!(| B!(vivisection)|;| B!(Government funding of vivisection)|)
B!(Government funding of vivisection)
Is this a valid argument? The typical way in which we determine whether an argument is valid is by asking whether there would be an inconsistency in affirming the premises but denying the conclusion. The above argument seems to display an inconsistency if we deny the conclusion whilst affirming the premises. If someone disapproves of vivisection and approves of the combination of disapproving both vivisection and government funding of vivisection then by not disapproving of the government funding of vivisection (by denying the conclusion) that person ‘ would fail to have a combination of attitudes of which he himself approves’ (Miller 2003: 61). The inconsistency arises then because someone is committed to approving of somebody who disapproves of both vivisection and the government funding of vivisection but himself failing to have such an attitude of disapproval of both disapproval of vivisection and government funding of vivisection by only disapproving of vivisection and not disapproving of the government funding of vivisection. It seems therefore that someone who holds (7) and (8) must hold (9) because if they don’t hold (9) then they are not disapproving what they approve others of disapproving of. As there seems to be an inconsistency in denying the conclusion whilst affirming the premises then the argument is valid. It is important to realise that the inconsistency generated is not a logical one but rather an attitudinal one. The non-cognitivist denies that (7),(8) and (9) are truth apt and given that two statements are only logically inconsistent if the truth of one entails the falsity of another (and vice versa) then the non-cognitivist cannot say that the denial of (9) cannot be logically inconsistent with (7) and (8). But nevertheless, according to the quasi-realist, the inconsistency generated is still a solution to the Frege-Geach problem. This is because the quasi-realist makes a distinction between the ‘ surface form of a discourse and the deep for of a discourse’ (Miller 2003: 60). The quasi-realist maintains that moral statements are superficially cognitive. That is, he maintains that moral statements appear to express descriptive mental states and therefore display statements which appear as if they describe some moral reality such as by the use of the predicate ‘ is wrong’ which implies that moral properties exist. But of course the quasi-realist denies that moral statements fundamentally express cognitive mental states but instead that they express non-descriptive mental attitudes. With this distinction made it is easy to see why the Frege-Geach problem is not problematic. For if moral statements are fundamentally expressions of non-cognitive mental states then whilst those statements cannot logically contradict each other they can result in a conflict of attitudes such as that given by the denial of (9) along with the assertions of (7) and (8). If the superficial cognitive nature of moral statements reflects the fundamental non-cognitive nature of moral statements then we might expect to find a clash of non-cognitive attitudes reflected in a cognitive form such as logical inconsistency. We think that the premises (7) and (8) do entail (9) because denying (9) results in an inconsistent set of attitudes because it clashes with (7) and (8). Likewise we find (1) and (2) do validly entail (3) because we find that they have superficially cognitive features which escapes the problem of equivocation found in the Frege-Geach problem. Therefore the quasi-realist can maintain both that moral statements are fundamentally non-cognitive whilst allowing cognitive features of moral statements. Quasi-realism is then easily summed up in the proposition that we express moral statements as if they were expressions of cognitive beliefs even though they are not.
The problem that the quasi-realist now faces is developing an account of the cognitive features of moral statements that is not inconsistent with the anti-realism of quasi-realism. One of the central tenets of moral realism is the idea that moral facts exist. That is, moral statements are true because they refer to moral properties that exist in the world. Quasi-realism denies that there are objective (not mind-dependent) moral properties (and thus facts) in the world. In fact, Blackburn gives an argument against moral realism on the basis that it is metaphysically extravagant. But if quasi-realism is anti-realist then how can moral statements even have the surface cognitive feature of truth-aptness? For if some moral statements have the superficial appearance of truth then given the
Correspondence theory of truth which states that a statement is true if and only if it correctly describes the world
then there must be some moral reality that makes such moral statements true. But this conflicts with anti-realism of quasi-realism. In order to address this issue Blackburn develops an alternative conception of truth that is broadly coherentist in nature. His strategy is to take the best possible set of attitudes (the best possible moral sensibility) and to declare a moral statement true as long as it is a member of this sensibility. The best possible moral sensibility is one ‘ which would result from taking all possible opportunities for improvement of attitude’ (Blackburn 1984: 198). At a minimum we might say that one moral sensibility is better than another moral sensibility if there are fewer inconsistent attitudes within that moral sensibility. Again, the notion of inconsistency here is not a logical one but merely inconsistent attitudes such as H!(vivisection) and B!(vivisection) . We would therefore take a moral statement such as (2) as being true as long as the non-cogntive mental attitude which it expresses (7) is a member of the best possible set of attitudes. The attitude itself is not true (because if it was then it would undermine the fundamentally non-cognitive nature of moral statements that quasi-realism affirms) but the moral statement that expresses it has the surface features of truth. The quasi-realist can therefore maintain his anti-realist metaphysics whilst maintaining surface cognitive features which allow him to escape the Frege-Geach problem. It might be objected
In conclusion, the quasi-realist account of moral statements is plausible. Quasi-realism is a non-cognitivist position that retains many of the beneficial parts of traditional non-cognitivism whilst overcoming the major problem that all such theories faced; The Frege-Geach problem. The problem for non-cognitivist theories was that such theories failed to confer validity on intuitively valid arguments because it gave different meanings to asserted and unasserted moral statements (thus showing valid arguments guilty of the fallacy of equivocation). Quasi-realism can successfully deal with this problem by interpreting valid moral arguments in terms of attitudes of approval and disapproval. An interpreted argument must be valid because otherwise inconsistent attitudes of approval arise. Further the quasi-realist argues that whilst the deep structure of moral statements is non-cognitive they have surface cognitive features which would allow the quasi-realist to escape the Frege-Geach problem. I finally argued that the quasi-realist can escape charges of inconsistency of its anti-realism and its support of cognitive features of moral statements by developing an alternative conception of truth.
